Ma T H, Xu J, Xia W, Jong X, Sun W, Lin G
Department of Biological Sciences, Western Illinois University, Macomb 61455.
Mutat Res. 1992 Nov 1;270(1):39-44. doi: 10.1016/0027-5107(92)90099-n.
The Tradescantia-micronucleus (Trad-MCN) bioassay is an efficient short-term test for genotoxicity of pollutants. In order to increase the efficiency and to standardize the micronucleus (MCN) scoring process, an automated scoring system was developed using the principle of image analysis in computer science. This assemblage is called the Tradescantia-micronucleus image analysis (Trad-MCNIA) system. The MCN frequencies scored by this system were compared with those scored by human observation for its proficiency. A set of low MCN frequency (around 5 MCN/100 tetrads) slides prepared from a control group, a set of medium MCN frequency (around 20 MCN/100 tetrads) slides prepared from sodium azide treated plant cuttings and a set of high MCN frequency (around 50 MCN/100 tetrads) slides prepared from X-ray treated materials were used for this study. In the low MCN frequency slides, the Trad-MCNIA system scored about the same value as human observation. In the medium and high frequency slides, MCN frequencies scored by the system were lower than those scored by human observers. This discrepancy was corrected by increasing the power of the objective of the microscope in the system. The MCN frequencies scored by the system attained 90% congruity with those scored by human observers after the correction. The scoring speed of the system was about 3.5 times as fast as that by human observers, and the data could be statistically analyzed immediately after the data scores were recorded. Further improvements can be made by upgrading the video camera and the computer speed.
紫露草微核(Trad-MCN)生物测定法是一种用于检测污染物遗传毒性的高效短期试验。为了提高效率并规范微核(MCN)评分过程,利用计算机科学中的图像分析原理开发了一种自动评分系统。这个组合被称为紫露草微核图像分析(Trad-MCNIA)系统。将该系统评分的MCN频率与人工观察评分的频率进行比较,以检验其准确性。本研究使用了一组由对照组制备的低MCN频率(约5个MCN/100个四分体)载玻片、一组由叠氮化钠处理过的植物插条制备的中等MCN频率(约20个MCN/100个四分体)载玻片和一组由X射线处理材料制备的高MCN频率(约50个MCN/100个四分体)载玻片。在低MCN频率载玻片中,Trad-MCNIA系统的评分与人工观察的评分大致相同。在中等和高频率载玻片中,该系统评分的MCN频率低于人工观察者评分的频率。通过增加系统中显微镜物镜的放大倍数来纠正这种差异。校正后,该系统评分的MCN频率与人工观察者评分的频率达到了90%的一致性。该系统的评分速度约为人工观察者的3.5倍,数据记录后可立即进行统计分析。通过升级摄像机和提高计算机速度可以进一步改进。